Estimating the Right-Tail Probability in Log-Gamma and Log-Weibull Distributions

• 발행기관: 한국자료분석학회
• 저자명: 강준호(Junho Kang) 이장춘(Jangchoon Lee)
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.9 No.6, 2677~2688쪽, 전체 12쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2007.12.30

영문 초록

We consider estimation of the right-tail probability in a log-gamma distribution and a log-Weibull distribution having each known shape parameter, especially we derive a UMVUE of the right-tail probability in the log-gamma and a log-Weibull distributions having each known shape parameter. And we derive densities of product, quotient, and the ratio of two independent log-gamma random variables. And we obtain the several facts through the numerical performance: (a) The log-Weibull density is skewed to the right side when alpha is greater than 1, but skewed to the left side when alpha is 1/2 and 1/3. (b) The kurtosis of the density is larger as alpha goes to a number(≥1/2). (c) The UMVUE performs better than the estimator tilde{R}(t) when t is larger time, and a proposed estimator tilde{R}(t) performs better than the UMVUE when t is small time in a sense of MSE.

목차

1. Introduction

2. Log-Gamma distribution

3. A Log-Weibull distribution

References

키워드

Log-gamma distribution Log-Weibull Right-Tail Probability